The numbers of reported cases of fistula between the portal venous system and adjacent organs has been gradually increasing recently. For instance, fistula has been reported between the portal venous system and the pancreas 12, biliary system 34, small intestine 567, and colon 89. However, fistula between the main portal vein and duodenum is extremely rare, with only three reported cases 567. Herein, we report an extremely rare case of a fistula between the main portal vein and duodenum after resection of extrahepatic bile duct due to hilar cholangiocarcinoma, and we discuss the strategy of duodenal peptic ulcer with massive bleeding.

A 68-year-old man with a diagnosis of anemia was admitted to Wakayama Medical University Hospital. He had a history of extended right hepatectomy for hilar cholangiocarcinoma (stage 2) 20 months earlier. On arrival, his blood pressure and pulse rate were 99/54 mm Hg and 101/min, respectively. Initial laboratory studies showed his hematocrit and hemoglobin were 25.8% and 8.2 g/dL. Nasogastric tube lavage revealed a material that looked like coffee grounds. Emergent gastrointestinal endoscopic examination was performed immediately and disclosed a giant peptic ulcer with active bleeding at the posterior wall of the duodenal bulbs (Figure 1a). Angiography was performed for the bleeding duodenal ulcer, which was not controlled by endoscopic hemostasis. Extravasation of contrast medium was not noted at the artery around the duodenum; however, transarterial embolization (TAE) was accomplished for the duodenal branch of the gastroduodenal artery. Massive hematemesis and anal bleeding occurred on the sixth day after the first TAE. The bleeding point of the ulcer was endoscopically treated with a clip (Figure 1b). Nevertheless, because bleeding from the duodenal ulcer occurred repeatedly, endoscopic treatment and TAE was performed, and endoscopic hemostasis was performed on the 10^th, 14^th, and 17^th day, and TAE was performed on the 10^th, 14^th, 19^th day. Finally, we embolized the common hepatic artery, the bilateral subphrenic artery, and the jejunal branch for hepaticojejunostomy. However, the patient died of sudden massive hematemesis on the 20^th hospital day. At autopsy, a peptic ulcer measuring 1.5 cm was present in the bulbus of the duodenum. The communication with the main portal vein and duodenal ulcer was manifested by insertion of a stick (Figure 2).

To our knowledge, only two cases have been reported that involved a penetrating ulcer of the duodenal bulbus extending into the main portal vein forming a fistula 67. Povoski et al. 7 reported a case with a fistula involving the portal vein and the duodenum at the site of a duodenal ulcer in a patient who had undergone previous extrahepatic bile duct resection and brachytherapy, which is similar to our case. They were successful in directly closing the fistula in the anterior wall of the portal vein after they took down the antecolic hepaticojejunostomies and divided the duodenal bulbus. Soares et al. 6 presented a case with a fistula between the duodenum and the portal vein caused by peptic ulcer with no history of previous surgery. They explained there was a small window superior to the pancreas and between the gastroduodenal artery and common bile duct where the portal vein and duodenum are separated by adipose tissue without pancreatic interposition (Additional file 1).

Bleeding is a serious complication of duodenal ulcers. Endoscopic hemostasis is the established first-line method for bleeding duodenal ulcers. The primary technical success rate is reportedly 90% in most studies 10. However, recurrent bleeding has an incidence of about 15% 11. If hemostasis is not achieved by endoscopic treatment, TAE or surgery may be required. However, controversy exists as to the safety and efficacy of these methods. Some authors have asserted that TAE is an effective method to stop massive bleeding from gastroduodenal ulcers in a high percentage of patients 1213. On the other hand, there are some reports that early elective surgery should be recommended in high-risk patients with bleeding duodenal bulb ulcer after failure of primary endoscopic treatment 14. The rates for emergency surgery are reportedly about 10% in general 15.

In the present case, because bleeding from the duodenal ulcer occurred repeatedly, endoscopic treatment and TAE were performed on all occasions without surgery, in expectation of the presence of numerous adhesions in the surrounding area of the mobilized duodenum.

There are few reports employing metallic stents in patients with malignant stenosis of the portal vein 16. In our case, if it had been revealed that the duodenal ulcer penetrated into the portal vein by extravascular extravasation of contrast medium in percutaneous transhepatic portography, the placement of the covered metallic stent may have been an effective modality with apprehension of patancy, before the repeated endoscopic and interventional radiologic therapy.

We should keep in mind that the main portal vein is exposed at the front of the hepatoduodenal ligament in cases with previous extrahepatic bile duct resection.

The author(s) declare that they have no competing interests.

HK designed the study and participated in writing process. KT was the main endoscopist managing the case. YN designed the study and collected the clinical data. MT helped to draft the manuscript.

KU participated in the editing process. HY revised the manuscript.

All authors read and approved the final manuscript.